Attention is currently required from: Nico Huber, Thomas Heijligen, Anastasia Klimchuk. Edward O'Callaghan has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/55295 )
Change subject: tests: Do not run a test if its driver is not built ......................................................................
Patch Set 5: Code-Review+2
(1 comment)
Commit Message:
https://review.coreboot.org/c/flashrom/+/55295/comment/9b7b4150_8dffdd9c PS5, Line 10: however config options can be disabled and in that case test should : not be run.
I'd even say it's better to return back to this later. Why switch […]
Agreed. This doesn't need to be resolved to the logical end point right now. The currently solution is perfectly reasonable for the number of tests we currently have.
Let's just resolve this and not be distracted by future problems here.
I totally agree with Nico's assessment about the weak symbol situation in coreboot!!! Can't stress that one enough and thanks for contrasting that here as it's a excellent case study why weak mostly sucks as well in the general case.
My only claim here is that, while I would think weak symbols would be a extremely exceptional to use, I suggest it is perhaps valid in this case. Only for the purposes of a test framework like cmocka that is designed around the weak symbol concept and that is to say it fits the paradigm and problem space. It is very unfortunately we do not have the features of Ada or Rust's traits at our disposal in C however if we did go the route of a clutter of `if (CONFIG(MEC1308))` I would prefer that it be all isolated into one object file in a very narrowly scoped way. Thanks C.
-
Edward O'Callaghan (Code Review)